메뉴
Delete pipeline data files to start fresh or re-run from a specific step. Supports full reset or partial reset from a given step onwards. Triggers on: "reset", "start over", "clear data", "reset pipeline", "reset from step", "delete data", "fresh start", "wipe data".
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
| name | reset |
| description | Delete pipeline data files to start fresh or re-run from a specific step. Supports full reset or partial reset from a given step onwards. Triggers on: "reset", "start over", "clear data", "reset pipeline", "reset from step", "delete data", "fresh start", "wipe data". |
Delete data files so the sourcing pipeline can be re-run from scratch or from a specific step.
| Step | Files produced | Path |
|---|---|---|
| 1 — source-ycombinator | yc_companies.json | data/raw/ |
| 2 — enrich-founders | enriched_founders.json | data/normalized/ |
| 3 — scrape-website | website_content.json | data/raw/ |
| 4 — find-latest-news | company_news.json | data/normalized/ |
| 5 — find-latest-fundraise | company_fundraises.json, sourced_companies.json | data/normalized/ |
Ask the VC which reset they want before doing anything:
What would you like to reset?
1. Full reset — delete all pipeline data (start completely fresh)
2. From step 2 — keep YC company list, re-run enrichment onwards
3. From step 3 — keep YC companies + founder enrichment, re-run website scraping onwards
4. From step 4 — keep everything through website scraping, re-run news onwards
5. From step 5 — keep everything through news, re-run fundraise lookup only
Reply with the step number or "full".
Wait for the VC to reply before deleting anything.
| Mode | Files to delete |
|---|---|
| Full reset (1) | All files listed above |
| From step 2 | enriched_founders.json, website_content.json, company_news.json, company_fundraises.json, sourced_companies.json |
| From step 3 | website_content.json, company_news.json, company_fundraises.json, sourced_companies.json |
| From step 4 | company_news.json, company_fundraises.json, sourced_companies.json |
| From step 5 | company_fundraises.json, sourced_companies.json |
After the VC confirms the reset mode, run the corresponding shell commands.
Full reset:
rm -f data/raw/yc_companies.json \
data/raw/website_content.json \
data/normalized/enriched_founders.json \
data/normalized/company_news.json \
data/normalized/company_fundraises.json \
data/normalized/sourced_companies.json
From step 2:
rm -f data/normalized/enriched_founders.json \
data/raw/website_content.json \
data/normalized/company_news.json \
data/normalized/company_fundraises.json \
data/normalized/sourced_companies.json
From step 3:
rm -f data/raw/website_content.json \
data/normalized/company_news.json \
data/normalized/company_fundraises.json \
data/normalized/sourced_companies.json
From step 4:
rm -f data/normalized/company_news.json \
data/normalized/company_fundraises.json \
data/normalized/sourced_companies.json
From step 5:
rm -f data/normalized/company_fundraises.json \
data/normalized/sourced_companies.json
Only delete files that actually exist — skip any that are already missing without error.
Confirm to the VC what was deleted and what step to run next:
Reset complete. Deleted:
- [list of deleted files]
Kept:
- [list of kept files]
Next step: run [step name] to continue.
Enrich founder profiles for each sourced YC company. Uses Exa people search to fill missing LinkedIn URLs and professional bios. Apify bulk-email-finder for any missing founder emails. Clearly labels all inferred data. Triggers on: "enrich founders", "find founder details", "get founder info", "step 2", "next step", "continue to enrich founders", "enrich people".
Convert any or all sourcing pipeline JSON files to CSV. Handles flattening of nested objects and arrays. Saves CSV files alongside the JSON files. Triggers on: "export csv", "convert to csv", "export data", "download csv", "export all", "get csv", "export companies", "export founders", "export results".
Find the latest fundraising round for each sourced YC company. Uses Exa and web search to surface Crunchbase data, TechCrunch funding articles, and SEC filings. Produces a structured funding history with round details. Triggers on: "find fundraise", "find funding", "find latest round", "check funding status", "step 5", "next step", "continue to find fundraise", "find investors", "find recent raises", "funding history".
Find the latest news, press coverage, and social signals for each sourced YC company. Uses Exa news search to surface recent TechCrunch articles, launch announcements, product launches, hiring news, and founder interviews. Triggers on: "find news", "get latest news", "find press coverage", "find articles", "step 4", "next step", "continue to find news", "get news about companies", "check news", "find recent coverage".
Scrape each YC company's website to extract product description, value proposition, target market, and key positioning. Uses Apify website-content-crawler for JS-heavy pages and Exa crawling for simpler sites. Triggers on: "scrape websites", "get website content", "scrape company sites", "step 3", "next step", "continue to scrape websites", "fetch website content", "scrape site".
Source YC startups from the Y Combinator directory using the Apify michael.g/y-combinator-scraper actor. Filters by batch, industry, region, or hiring status. Saves raw company and founder data. Triggers on: "source yc", "find yc companies", "scrape ycombinator", "source startups", "get yc batch", "start sourcing", "step 1", "find startups from yc", "source ycombinator".